中国东部夏季雨型与影响因子关系的年代际变化及预测模型新建

Construction and improvement of the prediction model, which is suitable for the climate warming, are in the particular command due to the shortage of the climate models and statistical diagnosis methods for the prediction of large scale rainbelt in China. This project aims at building the prediction model of the three patterns of summer rainfall in eastern China based on the investigations of the multi-time-scale and multi-space-scale changes, the abnormal characteristics and the mechanisms related to the summer rainbelt in eastern China under global warming. Three parts of work will be carried out in this project. First, the factors related to the three patterns of summer rainfall in eastern China will be classified according to the inter-annual and decadal changes of these factors. Then the analysis will be performed on the classified factors to figure out the reasons and mechanisms of the abnormal changes of summer-rain belt in eastern China. Second, the new possible predictors for the three patterns of summer rainfall in eastern China will be extracted based on the above analysis, and the impacts of these new predictors on the three patterns of summer rainfall in eastern China will be further investigated to expose the pathway of these factors in controlling the variations of the three patterns of summer rainfall in eastern China. Third, a new model for the prediction of the three patterns of summer rainfall in eastern China will be built based on the combination of the diagnostic methods of mathematical statistics with the numerical model technology. This model will have physical mechanism and higher accuracy, and is suitable for the background of global warming. In addition, some related scientific problems about the inter-annual and decadal variations of summer rain-belt in eastern China could be explained to some extent based on this model, which can promote the developments of the theory, technology and method of the summer rainfall prediction in eastern China.

目前气候模式和统计诊断方法对我国夏季大尺度雨带的跨季度预测能力仍然不高，构建和改进适合气候变暖背景下的预测模型迫在眉睫。本项目以中国东部地区夏季三类雨型的预测模型的新建为主要目标，研究全球变暖背景下中国东部地区夏季雨带多时、空尺度的变化和异常特征、成因和物理机制。首先，从探讨和检测中国东部地区夏季三类雨型与其主要影响因子之间年际关系和年代际的转变入手，对已发生转变和未发生转变的因子进行归类和分析，联系雨带发生转变和异常的年际和年代际异常的原因，分析物理机制；进而，探索和揭露中国东部地区夏季三类雨型新的可能的预测信号及其影响的途径；最后，将数理统计诊断方法和数值模式技术相结合，重新构建一个适合全球变暖背景、具有物理机制且有较高正确率的三类雨型预测模型，解释一些我国夏季雨型变异有关的科学问题，并提出关键技术和方法，完善我国东部地区夏季雨型预测的理论、技术和方法。

【项目的背景】夏季雨型预测是我国短期气候预测的重中之重。目前气候模式对我国夏季雨型的跨季度预测能力仍然不高，远远达不到实际需求；而统计诊断方法则面临雨型影响因子众多关系复杂，且全球变暖导致雨型与主要影响因子之间年际关系发生年代际变化这样复杂的困局。过去构建的三类雨型预测模型预测准确率明显下降。因此，构建和改进适合气候变暖背景下的预测模型迫在眉睫。.【主要研究内容】本项目以中国东部夏季雨型预测模型的新建为主要目标。首先，探讨和检测三类雨型与其主要影响因子之间年际关系和年代际的转变，探索和揭露三类雨型新的可能的预测信号及其影响的途径；最后，重新构建一个适合全球变暖背景、具有物理机制且有较高正确率的三类雨型预测模型。将预测模型进行独立样本回报和实际预测检验。此外，结合研究中发现的问题，重新审视了变暖背景下雨型的新特点，客观划分了东部季风区的四类雨型，进一步研究了四类雨型的影响环流系统、水汽输送和预测信号，改进了预测模型。.【重要结果】研究发现在全球变暖的背景下，ENSO、NPO等主要影响因子与中国夏季雨型之间的关系发生了显著地改变，对中国夏季降水的指示意义减弱，导致三类雨型早期预测概念模型的准确率明显下降。基于新的年代际气候背景，利用前冬欧亚型遥相关和太平洋/北美遥相关型，对中国东部夏季三类雨型预测概念模型进行了新建，其拟合效果较理想。并将三类雨型细化为四类雨型，四类雨型出现的时段具有显著的年代际特征，且四类雨型的影响环流系统和水汽输送存在明显的差异。发现冬春季印度洋海温异常及南印度洋副热带偶极子的位相是长江型和华南型的主要预测因子。进而将印度洋海温信号融合于雨型预测概念模型，改进了预测效果。.【关键数据及其科学意义】目前，新建的雨型预测模型已在国家气候中心季节预测业务中得到了应用，并在近三年实际预测中较好地预测了雨型，取得了良好的预测服务效果，为提高我国防灾减灾能力提供了技术支撑。